Comprehensive Understanding of the Zipingpu Reservoir to the Ms8.0 Wenchuan Earthquake

Abstract:

After the Wenchuan earthquake occurred, whether the big earthquake triggered by the storage of the Zipingpu Reservoir has attracted wide attention in international academic community. In addition to the qualitative discussion, many scholars also adopted the quantitative analysis methods to calculate the stress changes, but due to the different results, they draw very different conclusions. Here, we take the dispute of different teams in the quantitative calculation of Zipingpu reservoir as a starting point. In order to find out the key influence factors of quantitative calculation and know about the existing uncertainty elements during the numerical simulation, we analyze factors which may cause the differences. The preliminary results show that the calculation methods (analytical method or numerical method), dimension of models (2-D or 3-D), diffusion model, diffusion coefficient and focal mechanism are the main factors resulted in the differences, especially the diffusion coefficient of the fractured rock mass. The change of coulomb failure stress of the epicenter of Wenchuan earthquake attained from 2-D model is about 3 times of that of 3-D model. And it is not reasonable that only considering the fault permeability (assuming the permeability of rock mass as infinity) or only considering homogeneous isotropic rock mass permeability (ignoring the fault permeability). The different focal mechanisms also could dramatically affect the change of coulomb failure stress of the epicenter of Wenchuan earthquake, and the differences can research 2-7 times. And the differences the change of coulomb failure stress can reach several hundreds times, when selecting different diffusion coefficients. According to existing research that the magnitude of coulomb failure stress change is about several kPa, we could not rule out the possibility that the Zipingpu Reservoir may trigger the 2008 Wenchuan earthquake. However, for the background stress is not clear and coulomb failure stress change is too little, we also not sure there must be a connection between reservoir and earthquake. In future work, we should target on the basis of field survey and indoor experiment, improve the model and develop high performance simulation.